Showing 1–50 of 168 results for author: Jeon, E

FIESTA: Fourier-Based Semantic Augmentation with Uncertainty Guidance for Enhanced Domain Generalizability in Medical Image Segmentation

Authors: Kwanseok Oh, Eun** Jeon, Da-Woon Heo, Yooseung Shin, Heung-Il Suk

Abstract: Single-source domain generalization (SDG) in medical image segmentation (MIS) aims to generalize a model using data from only one source domain to segment data from an unseen target domain. Despite substantial advances in SDG with data augmentation, existing methods often fail to fully consider the details and uncertain areas prevalent in MIS, leading to mis-segmentation. This paper proposes a Fou… ▽ More Single-source domain generalization (SDG) in medical image segmentation (MIS) aims to generalize a model using data from only one source domain to segment data from an unseen target domain. Despite substantial advances in SDG with data augmentation, existing methods often fail to fully consider the details and uncertain areas prevalent in MIS, leading to mis-segmentation. This paper proposes a Fourier-based semantic augmentation method called FIESTA using uncertainty guidance to enhance the fundamental goals of MIS in an SDG context by manipulating the amplitude and phase components in the frequency domain. The proposed Fourier augmentative transformer addresses semantic amplitude modulation based on meaningful angular points to induce pertinent variations and harnesses the phase spectrum to ensure structural coherence. Moreover, FIESTA employs epistemic uncertainty to fine-tune the augmentation process, improving the ability of the model to adapt to diverse augmented data and concentrate on areas with higher ambiguity. Extensive experiments across three cross-domain scenarios demonstrate that FIESTA surpasses recent state-of-the-art SDG approaches in segmentation performance and significantly contributes to boosting the applicability of the model in medical imaging modalities. △ Less

Submitted 20 June, 2024; originally announced June 2024.

Comments: 40 pages, 7 figures, 5 tables

Exclusion of the Cosmological Triangle in Reactor-Based Search for Axion-Like Particles

Authors: Byung Ju Park, Jae ** Choi, Eunju Jeon, **yu Kim, Kyungwon Kim, Sung Hyun Kim, Sun Kee Kim, Yeongduk Kim, Young Ju Ko, Byoung-Cheol Koh, Chang Hyon Ha, Seo Hyun Lee, In Soo Lee, Hyunseok Lee, Hyun Su Lee, Jaison Lee, Yoomin Oh, Doo** Kim

Abstract: We report new constraints on axion-like particle (ALP) using data corresponding to a sodium iodine target exposure of 3063 kg$\cdot$days from the neutrino elastic scattering observation with NaI (NEON) experiment. A 16.7 kg of thallium-doped sodium iodide target was located 23.7 meters from a 2.8 GW thermal power nuclear reactor. We searched for ALPs produced by high-flux photons by comparing the… ▽ More We report new constraints on axion-like particle (ALP) using data corresponding to a sodium iodine target exposure of 3063 kg$\cdot$days from the neutrino elastic scattering observation with NaI (NEON) experiment. A 16.7 kg of thallium-doped sodium iodide target was located 23.7 meters from a 2.8 GW thermal power nuclear reactor. We searched for ALPs produced by high-flux photons by comparing the energy spectra of data collected during reactor-on (1596 kg$\cdot$days exposure) and reactor-off (1467 kg$\cdot$days exposure) periods. No signal consistent with ALP interaction was identified, allowing us to set exclusion limits at the 95% confidence level. Our limits cover previously unexplored regions for both photon couplings (${g_{aγ}}$) and electron couplings (${g_{ae}}$) for axion masses around 1 MeV/c$^2$. Notably, the NEON data excludes the unconstrained region identified by laboratory-based searches for photon couplings within the "cosmological triangle" for the first time. The observed 95\% confidence level limits reach as low as ${g_{aγ}}$ of 4.33$\times$ 10$^{-8}$ GeV$^{-1}$ and ${g_{ae}}$ of 1.10$\times$ 10$^{-9}$ for axion masses of 1.7 MeV/c$^2$ and 1.0 MeV/c$^2$, respectively. △ Less

Submitted 11 June, 2024; v1 submitted 10 June, 2024; originally announced June 2024.

Upgrade of NaI(Tl) crystal encapsulation for the NEON experiment

Authors: J. J. Choi, E. J. Jeon, J. Y. Kim, K. W. Kim, S. H. Kim, S. K. Kim, Y. D. Kim, Y. J. Ko, B. C. Koh, C. Ha, B. J. Park, S. H. Lee, I. S. Lee, H. Lee, H. S. Lee, J. Lee, Y. M. Oh

Abstract: The Neutrino Elastic-scattering Observation with NaI(Tl) experiment (NEON) aims to detect coherent elastic neutrino-nucleus scattering~(\cenns) in a NaI(Tl) crystal using reactor anti-electron neutrinos at the Hanbit nuclear power plant complex. A total of 13.3 kg of NaI(Tl) crystals were initially installed in December 2020 at the tendon gallery, 23.7$\pm$0.3\,m away from the reactor core, which… ▽ More The Neutrino Elastic-scattering Observation with NaI(Tl) experiment (NEON) aims to detect coherent elastic neutrino-nucleus scattering~(\cenns) in a NaI(Tl) crystal using reactor anti-electron neutrinos at the Hanbit nuclear power plant complex. A total of 13.3 kg of NaI(Tl) crystals were initially installed in December 2020 at the tendon gallery, 23.7$\pm$0.3\,m away from the reactor core, which operates at a thermal power of 2.8\,GW. Initial engineering operation was performed from May 2021 to March 2022 and observed unexpected photomultiplier-induced noise and a decreased light yield that were caused by leakage of liquid scintillator into the detector due to weakness of detector encapsulation. We upgraded the detector encapsulation design to prevent the leakage of the liquid scintillator. Meanwhile two small-sized detectors were replaced with larger ones resulting in a total mass of 16.7\,kg. With this new design implementation, the detector system has been operating stably since April 2022 for over a year without detector gain drop. In this paper, we present an improved crystal encapsulation design and stability of the NEON experiment. △ Less

Submitted 28 June, 2024; v1 submitted 2 April, 2024; originally announced April 2024.

Learning Decomposable and Debiased Representations via Attribute-Centric Information Bottlenecks

Authors: **yung Hong, Eun Som Jeon, Changhoon Kim, Keun Hee Park, Utkarsh Nath, Yezhou Yang, Pavan Turaga, Theodore P. Pavlic

Abstract: Biased attributes, spuriously correlated with target labels in a dataset, can problematically lead to neural networks that learn improper shortcuts for classifications and limit their capabilities for out-of-distribution (OOD) generalization. Although many debiasing approaches have been proposed to ensure correct predictions from biased datasets, few studies have considered learning latent embeddi… ▽ More Biased attributes, spuriously correlated with target labels in a dataset, can problematically lead to neural networks that learn improper shortcuts for classifications and limit their capabilities for out-of-distribution (OOD) generalization. Although many debiasing approaches have been proposed to ensure correct predictions from biased datasets, few studies have considered learning latent embedding consisting of intrinsic and biased attributes that contribute to improved performance and explain how the model pays attention to attributes. In this paper, we propose a novel debiasing framework, Debiasing Global Workspace, introducing attention-based information bottlenecks for learning compositional representations of attributes without defining specific bias types. Based on our observation that learning shape-centric representation helps robust performance on OOD datasets, we adopt those abilities to learn robust and generalizable representations of decomposable latent embeddings corresponding to intrinsic and biasing attributes. We conduct comprehensive evaluations on biased datasets, along with both quantitative and qualitative analyses, to showcase our approach's efficacy in attribute-centric representation learning and its ability to differentiate between intrinsic and bias-related features. △ Less

Submitted 21 March, 2024; originally announced March 2024.

Comments: 24 pages, 16 figures, 3 tables

Panda-70M: Captioning 70M Videos with Multiple Cross-Modality Teachers

Authors: Tsai-Shien Chen, Aliaksandr Siarohin, Willi Menapace, Ekaterina Deyneka, Hsiang-wei Chao, Byung Eun Jeon, Yuwei Fang, Hsin-Ying Lee, Jian Ren, Ming-Hsuan Yang, Sergey Tulyakov

Abstract: The quality of the data and annotation upper-bounds the quality of a downstream model. While there exist large text corpora and image-text pairs, high-quality video-text data is much harder to collect. First of all, manual labeling is more time-consuming, as it requires an annotator to watch an entire video. Second, videos have a temporal dimension, consisting of several scenes stacked together, a… ▽ More The quality of the data and annotation upper-bounds the quality of a downstream model. While there exist large text corpora and image-text pairs, high-quality video-text data is much harder to collect. First of all, manual labeling is more time-consuming, as it requires an annotator to watch an entire video. Second, videos have a temporal dimension, consisting of several scenes stacked together, and showing multiple actions. Accordingly, to establish a video dataset with high-quality captions, we propose an automatic approach leveraging multimodal inputs, such as textual video description, subtitles, and individual video frames. Specifically, we curate 3.8M high-resolution videos from the publicly available HD-VILA-100M dataset. We then split them into semantically consistent video clips, and apply multiple cross-modality teacher models to obtain captions for each video. Next, we finetune a retrieval model on a small subset where the best caption of each video is manually selected and then employ the model in the whole dataset to select the best caption as the annotation. In this way, we get 70M videos paired with high-quality text captions. We dub the dataset as Panda-70M. We show the value of the proposed dataset on three downstream tasks: video captioning, video and text retrieval, and text-driven video generation. The models trained on the proposed data score substantially better on the majority of metrics across all the tasks. △ Less

Submitted 29 February, 2024; originally announced February 2024.

Comments: CVPR 2024. Project Page: https://snap-research.github.io/Panda-70M

Waveform Simulation for Scintillation Characteristics of NaI(Tl) Crystal

Authors: J. J. Choi, C. Ha, E. J. Jeon, K. W. Kim, S. K. Kim, Y. D. Kim, Y. J. Ko, B. C. Koh, H. S. Lee, S. H. Lee, S. M. Lee, B. J. Park, G. H. Yu

Abstract: The lowering of the energy threshold in the NaI detector is crucial not only for comprehensive validation of DAMA/LIBRA but also for exploring new possibilities in the search for low-mass dark matter and observing coherent elastic scattering between neutrino and nucleus. Alongside hardware enhancements, extensive efforts have focused on refining event selection to discern noise, achieved through p… ▽ More The lowering of the energy threshold in the NaI detector is crucial not only for comprehensive validation of DAMA/LIBRA but also for exploring new possibilities in the search for low-mass dark matter and observing coherent elastic scattering between neutrino and nucleus. Alongside hardware enhancements, extensive efforts have focused on refining event selection to discern noise, achieved through parameter development and the application of machine learning. Acquiring pure, unbiased datasets is crucial in this endeavor, for which a waveform simulation was developed. The simulation data were compared with the experimental data using several pulse shape discrimination parameters to test its performance in describing the experimental data. Additionally, we present the outcomes of multi-variable machine learning trained with simulation data as a scintillation signal sample. The distributions of outcomes for experimental and simulation data show a good agreement. As an application of the waveform simulation, we validate the trigger efficiency alongside estimations derived from the minimally biased measurement data. △ Less

Submitted 17 June, 2024; v1 submitted 26 February, 2024; originally announced February 2024.

Journal ref: NIM A 1065, 169489 (2024)

νOscillation: a software package for computation and simulation of neutrino propagation and interaction

Authors: Seonghyeok Jang, Eunju Jeon, Eunil Won, Young Ju Ko, Kyungmin Lee

Abstract: The behavior of neutrinos is the only phenomenon that cannot be explained by the standard model of particle physics. Because of these mysterious neutrino interactions observed in nature, at present, there is growing interest in this field and ongoing or planned neutrino experiments are seeking solutions to this mystery very actively. The design of neutrino experiments and the analysis of neutrino… ▽ More The behavior of neutrinos is the only phenomenon that cannot be explained by the standard model of particle physics. Because of these mysterious neutrino interactions observed in nature, at present, there is growing interest in this field and ongoing or planned neutrino experiments are seeking solutions to this mystery very actively. The design of neutrino experiments and the analysis of neutrino data rely on precise computations of neutrino oscillations and scattering processes in general. Motivated by this, we developed a software package that calculates neutrino production and oscillation in nuclear reactors, neutrino-electron scattering of solar neutrinos, and the oscillation of neutrinos from radioactive isotopes for the search of sterile neutrinos. This software package is validated by reproducing the result of calculations and observations in other publications. We also demonstrate the feasibility of this package by calculating the sensitivity of a liquid scintillator detector, currently in planning, to the sterile neutrinos. This work is expected to be used in designs of future neutrino experiments. △ Less

Submitted 18 June, 2024; v1 submitted 23 January, 2024; originally announced January 2024.

Comments: 15 pages, 8 figures

Background study of the AMoRE-pilot experiment

Authors: A. Agrawal, V. V. Alenkov, P. Aryal, J. Beyer, B. Bhandari, R. S. Boiko, K. Boonin, O. Buzanov, C. R. Byeon, N. Chanthima, M. K. Cheoun, J. S. Choe, Seonho Choi, S. Choudhury, J. S. Chung, F. A. Danevich, M. Djamal, D. Drung, C. Enss, A. Fleischmann, A. M. Gangapshev, L. Gastaldo, Yu. M. Gavrilyuk, A. M. Gezhaev, O. Gileva , et al. (83 additional authors not shown)

Abstract: We report a study on the background of the Advanced Molybdenum-Based Rare process Experiment (AMoRE), a search for neutrinoless double beta decay (\znbb) of $^{100}$Mo. The pilot stage of the experiment was conducted using $\sim$1.9 kg of \CAMOO~ crystals at the Yangyang Underground Laboratory, South Korea, from 2015 to 2018. We compared the measured $β/γ$ energy spectra in three experimental conf… ▽ More We report a study on the background of the Advanced Molybdenum-Based Rare process Experiment (AMoRE), a search for neutrinoless double beta decay (\znbb) of $^{100}$Mo. The pilot stage of the experiment was conducted using $\sim$1.9 kg of \CAMOO~ crystals at the Yangyang Underground Laboratory, South Korea, from 2015 to 2018. We compared the measured $β/γ$ energy spectra in three experimental configurations with the results of Monte Carlo simulations and identified the background sources in each configuration. We replaced several detector components and enhanced the neutron shielding to lower the background level between configurations. A limit on the half-life of $0νββ$ decay of $^{100}$Mo was found at $T_{1/2}^{0ν} \ge 3.0\times 10^{23}$ years at 90\% confidence level, based on the measured background and its modeling. Further reduction of the background rate in the AMoRE-I and AMoRE-II are discussed. △ Less

Submitted 7 April, 2024; v1 submitted 15 January, 2024; originally announced January 2024.

Nonproportionality of NaI(Tl) Scintillation Detector for Dark Matter Search Experiments

Authors: S. M. Lee, G. Adhikari, N. Carlin, J. Y. Cho, J. J. Choi, S. Choi, A. C. Ezeribe, L. E. Fran. a, C. Ha, I. S. Hahn, S. J. Hollick, E. J. Jeon, H. W. Joo, W. G. Kang, M. Kauer, B. H. Kim, H. J. Kim, J. Kim, K. W. Kim, S. H. Kim, S. K. Kim, S. W. Kim, W. K. Kim, Y. D. Kim, Y. H. Kim , et al. (37 additional authors not shown)

Abstract: We present a comprehensive study of the nonproportionality of NaI(Tl) scintillation detectors within the context of dark matter search experiments. Our investigation, which integrates COSINE-100 data with supplementary $γ$ spectroscopy, measures light yields across diverse energy levels from full-energy $γ$ peaks produced by the decays of various isotopes. These $γ$ peaks of interest were produced… ▽ More We present a comprehensive study of the nonproportionality of NaI(Tl) scintillation detectors within the context of dark matter search experiments. Our investigation, which integrates COSINE-100 data with supplementary $γ$ spectroscopy, measures light yields across diverse energy levels from full-energy $γ$ peaks produced by the decays of various isotopes. These $γ$ peaks of interest were produced by decays supported by both long and short-lived isotopes. Analyzing peaks from decays supported only by short-lived isotopes presented a unique challenge due to their limited statistics and overlap** energies, which was overcome by long-term data collection and a time-dependent analysis. A key achievement is the direct measurement of the 0.87 keV light yield, resulting from the cascade following electron capture decay of $^{22}$Na from internal contamination. This measurement, previously accessible only indirectly, deepens our understanding of NaI(Tl) scintillator behavior in the region of interest for dark matter searches. This study holds substantial implications for background modeling and the interpretation of dark matter signals in NaI(Tl) experiments. △ Less

Submitted 10 May, 2024; v1 submitted 14 January, 2024; originally announced January 2024.

Comments: 12 pages, 7 figures

Journal ref: Eur. Phys. J. C 84 (2024) 484

Improving Bias Mitigation through Bias Experts in Natural Language Understanding

Authors: Eo** Jeon, Mingyu Lee, Juhyeong Park, Yeachan Kim, Wing-Lam Mok, SangKeun Lee

Abstract: Biases in the dataset often enable the model to achieve high performance on in-distribution data, while poorly performing on out-of-distribution data. To mitigate the detrimental effect of the bias on the networks, previous works have proposed debiasing methods that down-weight the biased examples identified by an auxiliary model, which is trained with explicit bias labels. However, finding a type… ▽ More Biases in the dataset often enable the model to achieve high performance on in-distribution data, while poorly performing on out-of-distribution data. To mitigate the detrimental effect of the bias on the networks, previous works have proposed debiasing methods that down-weight the biased examples identified by an auxiliary model, which is trained with explicit bias labels. However, finding a type of bias in datasets is a costly process. Therefore, recent studies have attempted to make the auxiliary model biased without the guidance (or annotation) of bias labels, by constraining the model's training environment or the capability of the model itself. Despite the promising debiasing results of recent works, the multi-class learning objective, which has been naively used to train the auxiliary model, may harm the bias mitigation effect due to its regularization effect and competitive nature across classes. As an alternative, we propose a new debiasing framework that introduces binary classifiers between the auxiliary model and the main model, coined bias experts. Specifically, each bias expert is trained on a binary classification task derived from the multi-class classification task via the One-vs-Rest approach. Experimental results demonstrate that our proposed strategy improves the bias identification ability of the auxiliary model. Consequently, our debiased model consistently outperforms the state-of-the-art on various challenge datasets. △ Less

Submitted 6 December, 2023; originally announced December 2023.

Comments: Accepted in EMNLP 2023 as a long paper

Alpha backgrounds in NaI(Tl) crystals of COSINE-100

Authors: G. Adhikari, N. Carlin, D. F. F. S. Cavalcante, J. Y. Cho, J. J. Choi, S. Choi, A. C. Ezeribe, L. E. Franca, C. Ha, I. S. Hahn, S. J. Hollick, E. J. Jeon, H. W. Joo, W. G. Kang, M. Kauer, B. H. Kim, H. J. Kim, J. Kim, K. W. Kim, S. H. Kim, S. K. Kim, S. W. Kim, W. K. Kim, Y. D. Kim, Y. H. Kim , et al. (38 additional authors not shown)

Abstract: COSINE-100 is a dark matter direct detection experiment with 106 kg NaI(Tl) as the target material. 210Pb and daughter isotopes are a dominant background in the WIMP region of interest and are detected via beta decay and alpha decay. Analysis of the alpha channel complements the background model as observed in the beta/gamma channel. We present the measurement of the quenching factors and Monte Ca… ▽ More COSINE-100 is a dark matter direct detection experiment with 106 kg NaI(Tl) as the target material. 210Pb and daughter isotopes are a dominant background in the WIMP region of interest and are detected via beta decay and alpha decay. Analysis of the alpha channel complements the background model as observed in the beta/gamma channel. We present the measurement of the quenching factors and Monte Carlo simulation results and activity quantification of the alpha decay components of the COSINE-100 NaI(Tl) crystals. The data strongly indicate that the alpha decays probabilistically undergo two possible quenching factors but require further investigation. The fitted results are consistent with independent measurements and improve the overall understanding of the COSINE-100 backgrounds. Furthermore, the half-life of 216Po has been measured to be 143.4 +/- 1.2 ms, which is consistent with and more precise than recent measurements. △ Less

Submitted 30 January, 2024; v1 submitted 8 November, 2023; originally announced November 2023.

Active search for a reactive target in thermal environments

Authors: Byeong Guk Go, Eui** Jeon, Yong Woon Kim

Abstract: We study a stochastic process where an active particle, modeled by a one-dimensional run-and-tumble particle, searches for a target with a finite absorption strength in thermal environments. Solving the Fokker-Planck equation for a uniform initial distribution, we analytically calculate the mean searching time (MST), the time for the active particle to be finally absorbed, and show that there exis… ▽ More We study a stochastic process where an active particle, modeled by a one-dimensional run-and-tumble particle, searches for a target with a finite absorption strength in thermal environments. Solving the Fokker-Planck equation for a uniform initial distribution, we analytically calculate the mean searching time (MST), the time for the active particle to be finally absorbed, and show that there exists an optimal self-propulsion velocity of the active particle at which MST is minimized. As the diffusion constant increases, the optimal velocity changes from a finite value to zero, which implies that a purely diffusive Brownian motion outperforms an active motion in terms of searching time. Depending on the absorption strength of the target, the transition of the optimal velocity becomes either continuous or discontinuous, which can be understood based on the Landau approach. In addition, we obtain the phase diagram indicating the passive-efficient and the active-efficient regions. Finally, the initial condition dependence of MST is presented in limiting cases. △ Less

Submitted 8 November, 2023; originally announced November 2023.

Searching for a partially absorbing target by a run-and-tumble particle in a confined space

Authors: Eui** Jeon, Byeongguk Go, Yong Woon Kim

Abstract: A random search of a partially absorbing target by a run-and-tumble particle in a confined one-dimensional space is investigated. We analytically obtain the mean searching time, which shows a non-monotonic behavior as a function of the self-propulsion speed of the active particle, indicating the existence of an optimal speed, when the absorption strength of the target is finite. In the limit of la… ▽ More A random search of a partially absorbing target by a run-and-tumble particle in a confined one-dimensional space is investigated. We analytically obtain the mean searching time, which shows a non-monotonic behavior as a function of the self-propulsion speed of the active particle, indicating the existence of an optimal speed, when the absorption strength of the target is finite. In the limit of large and small absorption strengths, respectively, asymptotes of the mean searching time and the optimal speed are found. We also demonstrate that the first-passage problem of a diffusive run-and-tumble particle in high dimensions can be mapped into a one-dimensional problem with a partially absorbing target. Finally, as a practical application exploiting the existence of the optimal speed, we propose a filtering device to extract active particles with a desired speed and evaluate how the resolution of the filtering device depends on the absorption strength. △ Less

Submitted 6 October, 2023; originally announced October 2023.

Comments: 12 pages, 3 figures

EAG-RS: A Novel Explainability-guided ROI-Selection Framework for ASD Diagnosis via Inter-regional Relation Learning

Authors: Wonsik Jung, Eun** Jeon, Eunsong Kang, Heung-Il Suk

Abstract: Deep learning models based on resting-state functional magnetic resonance imaging (rs-fMRI) have been widely used to diagnose brain diseases, particularly autism spectrum disorder (ASD). Existing studies have leveraged the functional connectivity (FC) of rs-fMRI, achieving notable classification performance. However, they have significant limitations, including the lack of adequate information whi… ▽ More Deep learning models based on resting-state functional magnetic resonance imaging (rs-fMRI) have been widely used to diagnose brain diseases, particularly autism spectrum disorder (ASD). Existing studies have leveraged the functional connectivity (FC) of rs-fMRI, achieving notable classification performance. However, they have significant limitations, including the lack of adequate information while using linear low-order FC as inputs to the model, not considering individual characteristics (i.e., different symptoms or varying stages of severity) among patients with ASD, and the non-explainability of the decision process. To cover these limitations, we propose a novel explainability-guided region of interest (ROI) selection (EAG-RS) framework that identifies non-linear high-order functional associations among brain regions by leveraging an explainable artificial intelligence technique and selects class-discriminative regions for brain disease identification. The proposed framework includes three steps: (i) inter-regional relation learning to estimate non-linear relations through random seed-based network masking, (ii) explainable connection-wise relevance score estimation to explore high-order relations between functional connections, and (iii) non-linear high-order FC-based diagnosis-informative ROI selection and classifier learning to identify ASD. We validated the effectiveness of our proposed method by conducting experiments using the Autism Brain Imaging Database Exchange (ABIDE) dataset, demonstrating that the proposed method outperforms other comparative methods in terms of various evaluation metrics. Furthermore, we qualitatively analyzed the selected ROIs and identified ASD subtypes linked to previous neuroscientific studies. △ Less

Submitted 5 October, 2023; originally announced October 2023.

Comments: 12 pages, 6 figures, 6 tables

Search for inelastic WIMP-iodine scattering with COSINE-100

Authors: G. Adhikari, N. Carlin, J. J. Choi, S. Choi, A. C. Ezeribe, L. E. Franca, C. Ha, I. S. Hahn, S. J. Hollick, E. J. Jeon, J. H. Jo, H. W. Joo, W. G. Kang, M. Kauer, B. H. Kim, H. J. Kim, J. Kim, K. W. Kim, S. H. Kim, S. K. Kim, W. K. Kim, Y. D. Kim, Y. H. Kim, Y. J. Ko, D. H. Lee , et al. (34 additional authors not shown)

Abstract: We report the results of a search for inelastic scattering of weakly interacting massive particles (WIMPs) off $^{127}$I nuclei using NaI(Tl) crystals with a data exposure of 97.7 kg$\cdot$years from the COSINE-100 experiment. The signature of inelastic WIMP-$^{127}$I scattering is a nuclear recoil accompanied by a 57.6 keV $γ$-ray from the prompt deexcitation, producing a more energetic signal co… ▽ More We report the results of a search for inelastic scattering of weakly interacting massive particles (WIMPs) off $^{127}$I nuclei using NaI(Tl) crystals with a data exposure of 97.7 kg$\cdot$years from the COSINE-100 experiment. The signature of inelastic WIMP-$^{127}$I scattering is a nuclear recoil accompanied by a 57.6 keV $γ$-ray from the prompt deexcitation, producing a more energetic signal compared to the typical WIMP nuclear recoil signal. We found no evidence for this inelastic scattering signature and set a 90 $\%$ confidence level upper limit on the WIMP-proton spin-dependent, inelastic scattering cross section of $1.2 \times 10^{-37} {\rm cm^{2}}$ at the WIMP mass 500 ${\rm GeV/c^{2}}$. △ Less

Submitted 30 October, 2023; v1 submitted 19 July, 2023; originally announced July 2023.

Comments: 8 pages, 5 figures. arXiv admin note: text overlap with arXiv:2104.03537

Journal ref: Phys. Rev. D 108, 092006 (2023)

Search for Boosted Dark Matter in COSINE-100

Authors: G. Adhikari, N. Carlin, J. J. Choi, S. Choi, A. C. Ezeribe, L. E. Franca, C. Ha, I. S. Hahn, S. J. Hollick, E. J. Jeon, J. H. Jo, H. W. Joo, W. G. Kang, M. Kauer, B. H. Kim, H. J. Kim, J. Kim, K. W. Kim, S. H. Kim, S. K. Kim, W. K. Kim, Y. D. Kim, Y. H. Kim, Y. J. Ko, D. H. Lee , et al. (34 additional authors not shown)

Abstract: We search for energetic electron recoil signals induced by boosted dark matter (BDM) from the galactic center using the COSINE-100 array of NaI(Tl) crystal detectors at the Yangyang Underground Laboratory. The signal would be an excess of events with energies above 4 MeV over the well-understood background. Because no excess of events are observed in a 97.7 kg$\cdot$years exposure, we set limits o… ▽ More We search for energetic electron recoil signals induced by boosted dark matter (BDM) from the galactic center using the COSINE-100 array of NaI(Tl) crystal detectors at the Yangyang Underground Laboratory. The signal would be an excess of events with energies above 4 MeV over the well-understood background. Because no excess of events are observed in a 97.7 kg$\cdot$years exposure, we set limits on BDM interactions under a variety of hypotheses. Notably, we explored the dark photon parameter space, leading to competitive limits compared to direct dark photon search experiments, particularly for dark photon masses below 4\,MeV and considering the invisible decay mode. Furthermore, by comparing our results with a previous BDM search conducted by the Super-Kamionkande experiment, we found that the COSINE-100 detector has advantages in searching for low-mass dark matter. This analysis demonstrates the potential of the COSINE-100 detector to search for MeV electron recoil signals produced by the dark sector particle interactions. △ Less

Submitted 30 October, 2023; v1 submitted 31 May, 2023; originally announced June 2023.

Comments: 7 pages, 4 figures

Journal ref: Phys. Rev. Lett. 131, 201802 (2023)

Search for bosonic super-weakly interacting massive particles at COSINE-100

Authors: G. Adhikari, N. Carlin, J. J. Choi, S. Choi, A. C. Ezeribe, L. E. Franca, C. Ha, I. S. Hahn, S. J. Hollick, E. J. Jeon, J. H. Jo, H. W. Joo, W. G. Kang, M. Kauer, B. H. Kim, H. J. Kim, J. Kim, K. W. Kim, S. H. Kim, S. K. Kim, W. K. Kim, Y. D. Kim, Y. H. Kim, Y. J. Ko, D. H. Lee , et al. (34 additional authors not shown)

Abstract: We present results of a search for bosonic super-weakly interacting massive particles (BSW) as keV scale dark matter candidates that is based on an exposure of 97.7 kg$\cdot$year from the COSINE experiment. In this search, we employ, for the first time, Compton-like as well as absorption processes for pseudoscalar and vector BSWs. No evidence for BSWs is found in the mass range from 10… ▽ More We present results of a search for bosonic super-weakly interacting massive particles (BSW) as keV scale dark matter candidates that is based on an exposure of 97.7 kg$\cdot$year from the COSINE experiment. In this search, we employ, for the first time, Compton-like as well as absorption processes for pseudoscalar and vector BSWs. No evidence for BSWs is found in the mass range from 10 $\mathrm{keV/c}^2$ to 1 $\mathrm{MeV/c}^2$, and we present the exclusion limits on the dimensionless coupling constants to electrons $g_{ae}$ for pseudoscalar and $κ$ for vector BSWs at 90% confidence level. Our results show that these limits are improved by including the Compton-like process in masses of BSW, above $\mathcal{O}(100\,\mathrm{keV/c}^2)$. △ Less

Submitted 27 August, 2023; v1 submitted 3 April, 2023; originally announced April 2023.

Journal ref: Phys. Rev. D 108 (2023) L041301

Leveraging Angular Distributions for Improved Knowledge Distillation

Authors: Eun Som Jeon, Hongjun Choi, Ankita Shukla, Pavan Turaga

Abstract: Knowledge distillation as a broad class of methods has led to the development of lightweight and memory efficient models, using a pre-trained model with a large capacity (teacher network) to train a smaller model (student network). Recently, additional variations for knowledge distillation, utilizing activation maps of intermediate layers as the source of knowledge, have been studied. Generally, i… ▽ More Knowledge distillation as a broad class of methods has led to the development of lightweight and memory efficient models, using a pre-trained model with a large capacity (teacher network) to train a smaller model (student network). Recently, additional variations for knowledge distillation, utilizing activation maps of intermediate layers as the source of knowledge, have been studied. Generally, in computer vision applications, it is seen that the feature activation learned by a higher capacity model contains richer knowledge, highlighting complete objects while focusing less on the background. Based on this observation, we leverage the dual ability of the teacher to accurately distinguish between positive (relevant to the target object) and negative (irrelevant) areas. We propose a new loss function for distillation, called angular margin-based distillation (AMD) loss. AMD loss uses the angular distance between positive and negative features by projecting them onto a hypersphere, motivated by the near angular distributions seen in many feature extractors. Then, we create a more attentive feature that is angularly distributed on the hypersphere by introducing an angular margin to the positive feature. Transferring such knowledge from the teacher network enables the student model to harness the higher discrimination of positive and negative features for the teacher, thus distilling superior student models. The proposed method is evaluated for various student-teacher network pairs on four public datasets. Furthermore, we show that the proposed method has advantages in compatibility with other learning techniques, such as using fine-grained features, augmentation, and other distillation methods. △ Less

Submitted 27 February, 2023; originally announced February 2023.

Comments: Neurocomputing, Volume 518, 21 January 2023, Pages 466-481

Journal ref: Neurocomputing, Volume 518, 2023, Pages 466-481

Search for solar bosonic dark matter annual modulation with COSINE-100

Authors: G. Adhikari, N. Carlin, J. J. Choi, S. Choi, A. C. Ezeribe, L. E. França, C. Ha, I. S. Hahn, S. J. Hollick, E. J. Jeon, J. H. Jo, H. W. Joo, W. G. Kang, M. Kauer, B. H. Kim, H. J. Kim, J. Kim, K. W. Kim, S. H. Kim, S. K. Kim, W. K. Kim, Y. D. Kim, Y. H. Kim, Y. J. Ko, D. H. Lee , et al. (34 additional authors not shown)

Abstract: We present results from a search for solar bosonic dark matter using the annual modulation method with the COSINE-100 experiment. The results were interpreted considering three dark sector bosons models: solar dark photon; DFSZ and KSVZ solar axion; and Kaluza-Klein solar axion. No modulation signal that is compatible with the expected from the models was found from a data-set of 2.82 yr, using 61… ▽ More We present results from a search for solar bosonic dark matter using the annual modulation method with the COSINE-100 experiment. The results were interpreted considering three dark sector bosons models: solar dark photon; DFSZ and KSVZ solar axion; and Kaluza-Klein solar axion. No modulation signal that is compatible with the expected from the models was found from a data-set of 2.82 yr, using 61.3 kg of NaI(Tl) crystals. Therefore, we set a 90$\%$ confidence level upper limits for each of the three models studied. For the solar dark photon model, the most stringent mixing parameter upper limit is $1.61 \times 10^{-14}$ for dark photons with a mass of 215 eV. For the DFSZ and KSVZ solar axion, and the Kaluza-Klein axion models, the upper limits exclude axion-electron couplings, $g_{ae}$, above $1.61 \times 10^{-11}$ for axion mass below 0.2 keV; and axion-photon couplings, $g_{aγγ}$, above $1.83 \times 10^{-11}$ GeV$^{-1}$ for an axion number density of $4.07 \times 10^{13}$ cm$^{-3}$. This is the first experimental search for solar dark photons and DFSZ and KSVZ solar axions using the annual modulation method. The lower background, higher light yield and reduced threshold of NaI(Tl) crystals of the future COSINE-200 experiment are expected to enhance the sensitivity of the analysis shown in this paper. We show the sensitivities for the three models studied, considering the same search method with COSINE-200. △ Less

Submitted 20 February, 2023; originally announced February 2023.

Comments: 13 pages, 16 figures

Performance of an ultra-pure NaI(Tl) detector produced by an indigenously-developed purification method and crystal growth for the COSINE-200 experiment

Authors: Hyun Seok Lee, Byung Ju Park, Jae ** Choi, Olga Gileva, Chang Hyon Ha, Alain Iltis, Eun Ju Jeon, Dae Yeon Kim, Kyung Won Kim, Sung Hyun Kim, Sun Kee Kim, Yeong Duk Kim, Young Ju Ko, Cheol Ho Lee, Hyun Su Lee, In Soo Lee, Moo Hyun Lee, Se ** Ra, Ju Kyung Son, Keon Ah Shin

Abstract: The COSINE-100 experiment has been operating with 106 kg of low-background NaI(Tl) detectors to test the results from the DAMA/LIBRA experiment, which claims to have observed dark matter. However, since the background of the NaI(Tl) crystals used in the COSINE-100 experiment is 2-3 times higher than that in the DAMA detectors, no conclusion regarding the claimed observation from the DAMA/LIBRA exp… ▽ More The COSINE-100 experiment has been operating with 106 kg of low-background NaI(Tl) detectors to test the results from the DAMA/LIBRA experiment, which claims to have observed dark matter. However, since the background of the NaI(Tl) crystals used in the COSINE-100 experiment is 2-3 times higher than that in the DAMA detectors, no conclusion regarding the claimed observation from the DAMA/LIBRA experiment could be reached. Therefore, we plan to upgrade the current COSINE-100 experiment to the next phase, COSINE-200, by using ultra-low background NaI(Tl) detectors. The basic principle was already proved with the commercially available Astro-grade NaI powder from Sigma-Aldrich company. However, we have developed a mass production process of ultra-pure NaI powder at the Center for Underground Physics (CUP) of the Institute for Basic Science (IBS), Korea, using the direct purification of the raw NaI powder. We plan to produce more than 1,000 kg of ultra-pure powder for the COSINE200 experiment. With our crystal grower installed at CUP, we have successfully grown a low-background crystal using our purification technique for the NaI powder. We have assembled a low-background NaI(Tl) detector. In this article, we report the performance of this ultra-pure NaI(Tl) crystal detector produced at IBS, Korea. △ Less

Submitted 12 January, 2023; originally announced January 2023.

Journal ref: Frontiers in Physics 11 2023

1st Workshop on Maritime Computer Vision (MaCVi) 2023: Challenge Results

Authors: Benjamin Kiefer, Matej Kristan, Janez Perš, Lojze Žust, Fabio Poiesi, Fabio Augusto de Alcantara Andrade, Alexandre Bernardino, Matthew Dawkins, Jenni Raitoharju, Yitong Quan, Adem Atmaca, Timon Höfer, Qiming Zhang, Yufei Xu, **g Zhang, Dacheng Tao, Lars Sommer, Raphael Spraul, Hangyue Zhao, Hongpu Zhang, Yanyun Zhao, Jan Lukas Augustin, Eui-ik Jeon, Impyeong Lee, Luca Zedda , et al. (48 additional authors not shown)

Abstract: The 1$^{\text{st}}$ Workshop on Maritime Computer Vision (MaCVi) 2023 focused on maritime computer vision for Unmanned Aerial Vehicles (UAV) and Unmanned Surface Vehicle (USV), and organized several subchallenges in this domain: (i) UAV-based Maritime Object Detection, (ii) UAV-based Maritime Object Tracking, (iii) USV-based Maritime Obstacle Segmentation and (iv) USV-based Maritime Obstacle Detec… ▽ More The 1$^{\text{st}}$ Workshop on Maritime Computer Vision (MaCVi) 2023 focused on maritime computer vision for Unmanned Aerial Vehicles (UAV) and Unmanned Surface Vehicle (USV), and organized several subchallenges in this domain: (i) UAV-based Maritime Object Detection, (ii) UAV-based Maritime Object Tracking, (iii) USV-based Maritime Obstacle Segmentation and (iv) USV-based Maritime Obstacle Detection. The subchallenges were based on the SeaDronesSee and MODS benchmarks. This report summarizes the main findings of the individual subchallenges and introduces a new benchmark, called SeaDronesSee Object Detection v2, which extends the previous benchmark by including more classes and footage. We provide statistical and qualitative analyses, and assess trends in the best-performing methodologies of over 130 submissions. The methods are summarized in the appendix. The datasets, evaluation code and the leaderboard are publicly available at https://seadronessee.cs.uni-tuebingen.de/macvi. △ Less

Submitted 28 November, 2022; v1 submitted 24 November, 2022; originally announced November 2022.

Comments: MaCVi 2023 was part of WACV 2023. This report (38 pages) discusses the competition as part of MaCVi

Pulse shape discrimination using a convolutional neural network for organic liquid scintillator signals

Authors: K. Y. Jung, B. Y. Han, E. J. Jeon, Y. Jeong, H. S. Jo, J. Y. Kim, J. G. Kim, Y. D. Kim, Y. J. Ko, M. H. Lee, J. Lee, C. S. Moon, Y. M. Oh, H. K. Park, S. H. Seo, D. W. Seol, K. Siyeon, G. M. Sun, Y. S. Yoon, I. Yu

Abstract: A convolutional neural network (CNN) architecture is developed to improve the pulse shape discrimination (PSD) power of the gadolinium-loaded organic liquid scintillation detector to reduce the fast neutron background in the inverse beta decay candidate events of the NEOS-II data. A power spectrum of an event is constructed using a fast Fourier transform of the time domain raw waveforms and put in… ▽ More A convolutional neural network (CNN) architecture is developed to improve the pulse shape discrimination (PSD) power of the gadolinium-loaded organic liquid scintillation detector to reduce the fast neutron background in the inverse beta decay candidate events of the NEOS-II data. A power spectrum of an event is constructed using a fast Fourier transform of the time domain raw waveforms and put into CNN. An early data set is evaluated by CNN after it is trained using low energy $β$ and $α$ events. The signal-to-background ratio averaged over 1-10 MeV visible energy range is enhanced by more than 20% in the result of the CNN method compared to that of an existing conventional PSD method, and the improvement is even higher in the low energy region. △ Less

Submitted 15 January, 2023; v1 submitted 14 November, 2022; originally announced November 2022.

Comments: 16 pages, 10 figures

Understanding the Role of Mixup in Knowledge Distillation: An Empirical Study

Authors: Hongjun Choi, Eun Som Jeon, Ankita Shukla, Pavan Turaga

Abstract: Mixup is a popular data augmentation technique based on creating new samples by linear interpolation between two given data samples, to improve both the generalization and robustness of the trained model. Knowledge distillation (KD), on the other hand, is widely used for model compression and transfer learning, which involves using a larger network's implicit knowledge to guide the learning of a s… ▽ More Mixup is a popular data augmentation technique based on creating new samples by linear interpolation between two given data samples, to improve both the generalization and robustness of the trained model. Knowledge distillation (KD), on the other hand, is widely used for model compression and transfer learning, which involves using a larger network's implicit knowledge to guide the learning of a smaller network. At first glance, these two techniques seem very different, however, we found that "smoothness" is the connecting link between the two and is also a crucial attribute in understanding KD's interplay with mixup. Although many mixup variants and distillation methods have been proposed, much remains to be understood regarding the role of a mixup in knowledge distillation. In this paper, we present a detailed empirical study on various important dimensions of compatibility between mixup and knowledge distillation. We also scrutinize the behavior of the networks trained with a mixup in the light of knowledge distillation through extensive analysis, visualizations, and comprehensive experiments on image classification. Finally, based on our findings, we suggest improved strategies to guide the student network to enhance its effectiveness. Additionally, the findings of this study provide insightful suggestions to researchers and practitioners that commonly use techniques from KD. Our code is available at https://github.com/hchoi71/MIX-KD. △ Less

Submitted 8 November, 2022; v1 submitted 7 November, 2022; originally announced November 2022.

Comments: To be presented at WACV 2023

Status and performance of the AMoRE-I experiment on neutrinoless double beta decay

Authors: H. B. Kim, D. H. Ha, E. J. Jeon, J. A. Jeon, H. S. Jo, C. S. Kang, W. G. Kang, H. S. Kim, S. C. Kim, S. G. Kim, S. K. Kim, S. R. Kim, W. T. Kim, Y. D. Kim, Y. H. Kim, D. H. Kwon, E. S. Lee, H. J. Lee, H. S. Lee, J. S. Lee, M. H. Lee, S. W. Lee, Y. C. Lee, D. S. Leonard, H. S. Lim , et al. (10 additional authors not shown)

Abstract: AMoRE is an international project to search for the neutrinoless double beta decay of $^{100}$Mo using a detection technology consisting of magnetic microcalorimeters (MMCs) and molybdenum-based scintillating crystals. Data collection has begun for the current AMORE-I phase of the project, an upgrade from the previous pilot phase. AMoRE-I employs thirteen $^\mathrm{48depl.}$Ca$^{100}$MoO$_4$ cryst… ▽ More AMoRE is an international project to search for the neutrinoless double beta decay of $^{100}$Mo using a detection technology consisting of magnetic microcalorimeters (MMCs) and molybdenum-based scintillating crystals. Data collection has begun for the current AMORE-I phase of the project, an upgrade from the previous pilot phase. AMoRE-I employs thirteen $^\mathrm{48depl.}$Ca$^{100}$MoO$_4$ crystals and five Li$_2$$^{100}$MoO$_4$ crystals for a total crystal mass of 6.2 kg. Each detector module contains a scintillating crystal with two MMC channels for heat and light detection. We report the present status of the experiment and the performance of the detector modules. △ Less

Submitted 5 November, 2022; originally announced November 2022.

Comments: 8 pages, 4 figures, published in Journal of Low Temperature Physics (2022)

An induced annual modulation signature in COSINE-100 data by DAMA/LIBRA's analysis method

Authors: G. Adhikari, N. Carlin, J. J. Choi, S. Choi, A. C. Ezeribe, L. E. Franca, C. Ha, I. S. Hahn, S. J. Hollick, E. J. Jeon, J. H. Jo, H. W. Joo, W. G. Kang, M. Kauer, B. H. Kim, H. J. Kim, J. Kim, K. W. Kim, S. H. Kim, S. K. Kim, W. K. Kim, Y. D. Kim, Y. H. Kim, Y. J. Ko, D. H. Lee , et al. (32 additional authors not shown)

Abstract: The DAMA/LIBRA collaboration has reported the observation of an annual modulation in the event rate that has been attributed to dark matter interactions over the last two decades. However, even though tremendous efforts to detect similar dark matter interactions were pursued, no definitive evidence has been observed to corroborate the DAMA/LIBRA signal. Many studies assuming various dark matter mo… ▽ More The DAMA/LIBRA collaboration has reported the observation of an annual modulation in the event rate that has been attributed to dark matter interactions over the last two decades. However, even though tremendous efforts to detect similar dark matter interactions were pursued, no definitive evidence has been observed to corroborate the DAMA/LIBRA signal. Many studies assuming various dark matter models have attempted to reconcile DAMA/LIBRA's modulation signals and null results from other experiments, however no clear conclusion can be drawn. Apart from the dark matter hypothesis, several studies have examined the possibility that the modulation is induced by variations in their detector's environment or their specific analysis methods. In particular, a recent study presents a possible cause of the annual modulation from an analysis method adopted by the DAMA/LIBRA experiment in which the observed annual modulation could be reproduced by a slowly varying time-dependent background. Here, we study the COSINE-100 data using an analysis method similar to the one adopted by the DAMA/LIBRA experiment and observe a significant annual modulation, although the modulation phase is almost opposite to that of the DAMA/LIBRA data. Assuming the same background composition for COSINE-100 and DAMA/LIBRA, simulated experiments for the DAMA/LIBRA without dark matter signals also provide significant annual modulation with an amplitude similar to DAMA/LIBRA with opposite phase. Even though this observation does not explain the DAMA/LIBRA's results directly, this interesting phenomenon motivates deeper studies of the time-dependent DAMA/LIBRA background data. △ Less

Submitted 10 August, 2022; originally announced August 2022.

Journal ref: Sci. Rep. 13, 4676 (2023)

Exploring coherent elastic neutrino-nucleus scattering using reactor electron antineutrinos in the NEON experiment

Authors: J. J. Choi, E. J. Jeon, J. Y. Kim, K. W. Kim, S. H. Kim, S. K. Kim, Y. D. Kim, Y. J. Ko, B. C. Koh, C. Ha, B. J. Park, S. H. Lee, I. S. Lee, H. Lee, H. S. Lee, J. Lee, Y. M. Oh

Abstract: Neutrino elastic scattering observation with NaI (NEON) is an experiment designed to detect neutrino-nucleus coherent scattering using reactor electron antineutrinos. NEON is based on an array of six NaI(Tl) crystals with a total mass of 13.3 kg, located at the tendon gallery that is 23.7 m away from a reactor core with a thermal power of 2.8 GW in the Hanbit nuclear power complex. The installatio… ▽ More Neutrino elastic scattering observation with NaI (NEON) is an experiment designed to detect neutrino-nucleus coherent scattering using reactor electron antineutrinos. NEON is based on an array of six NaI(Tl) crystals with a total mass of 13.3 kg, located at the tendon gallery that is 23.7 m away from a reactor core with a thermal power of 2.8 GW in the Hanbit nuclear power complex. The installation of the NEON detector was completed in December 2020, and since May 2021, the detector has acquired data at full reactor power. Based on the observed light yields of the NaI crystals of approximately 22, number of photoelectrons per unit keV electron-equivalent energy (keVee), and 6 counts/kg/keV/day background level at 2-6 keVee energy, coherent elastic neutrino-nucleus scattering observation sensitivity is evaluated as more than 3$σ$ assuming one-year reactor-on and 100 days reactor-off data, 0.2 keVee energy threshold, and 7 counts/keV/kg/day background in the signal region of 0.2-0.5 keVee. This paper describes the design of the NEON detector, including the shielding arrangement, configuration of NaI(Tl) crystals, and associated operating systems. The initial performance and associated sensitivity of the experiment are also presented. △ Less

Submitted 20 December, 2022; v1 submitted 8 April, 2022; originally announced April 2022.

Comments: 21 pages 27 figures

A Style-aware Discriminator for Controllable Image Translation

Authors: Kunhee Kim, Sanghun Park, Eunyeong Jeon, Taehun Kim, Dai** Kim

Abstract: Current image-to-image translations do not control the output domain beyond the classes used during training, nor do they interpolate between different domains well, leading to implausible results. This limitation largely arises because labels do not consider the semantic distance. To mitigate such problems, we propose a style-aware discriminator that acts as a critic as well as a style encoder to… ▽ More Current image-to-image translations do not control the output domain beyond the classes used during training, nor do they interpolate between different domains well, leading to implausible results. This limitation largely arises because labels do not consider the semantic distance. To mitigate such problems, we propose a style-aware discriminator that acts as a critic as well as a style encoder to provide conditions. The style-aware discriminator learns a controllable style space using prototype-based self-supervised learning and simultaneously guides the generator. Experiments on multiple datasets verify that the proposed model outperforms current state-of-the-art image-to-image translation methods. In contrast with current methods, the proposed approach supports various applications, including style interpolation, content transplantation, and local image translation. △ Less

Submitted 29 March, 2022; originally announced March 2022.

Comments: 2022 CVPR

TransSleep: Transitioning-aware Attention-based Deep Neural Network for Sleep Staging

Authors: Jauen Phyo, Wonjun Ko, Eun** Jeon, Heung-Il Suk

Abstract: Sleep staging is essential for sleep assessment and plays a vital role as a health indicator. Many recent studies have devised various machine learning as well as deep learning architectures for sleep staging. However, two key challenges hinder the practical use of these architectures: effectively capturing salient waveforms in sleep signals and correctly classifying confusing stages in transition… ▽ More Sleep staging is essential for sleep assessment and plays a vital role as a health indicator. Many recent studies have devised various machine learning as well as deep learning architectures for sleep staging. However, two key challenges hinder the practical use of these architectures: effectively capturing salient waveforms in sleep signals and correctly classifying confusing stages in transitioning epochs. In this study, we propose a novel deep neural network structure, TransSleep, that captures distinctive local temporal patterns and distinguishes confusing stages using two auxiliary tasks. In particular, TransSleep adopts an attention-based multi-scale feature extractor module to capture salient waveforms; a stage-confusion estimator module with a novel auxiliary task, epoch-level stage classification, to estimate confidence scores for identifying confusing stages; and a context encoder module with the other novel auxiliary task, stage-transition detection, to represent contextual relationships across neighboring epochs. Results show that TransSleep achieves promising performance in automatic sleep staging. The validity of TransSleep is demonstrated by its state-of-the-art performance on two publicly available datasets, Sleep-EDF and MASS. Furthermore, we performed ablations to analyze our results from different perspectives. Based on our overall results, we believe that TransSleep has immense potential to provide new insights into deep learning-based sleep staging. △ Less

Submitted 22 March, 2022; originally announced March 2022.

Comments: 13 pages, 9 figures

Role of Data Augmentation Strategies in Knowledge Distillation for Wearable Sensor Data

Authors: Eun Som Jeon, Anirudh Som, Ankita Shukla, Kristina Hasanaj, Matthew P. Buman, Pavan Turaga

Abstract: Deep neural networks are parametrized by several thousands or millions of parameters, and have shown tremendous success in many classification problems. However, the large number of parameters makes it difficult to integrate these models into edge devices such as smartphones and wearable devices. To address this problem, knowledge distillation (KD) has been widely employed, that uses a pre-trained… ▽ More Deep neural networks are parametrized by several thousands or millions of parameters, and have shown tremendous success in many classification problems. However, the large number of parameters makes it difficult to integrate these models into edge devices such as smartphones and wearable devices. To address this problem, knowledge distillation (KD) has been widely employed, that uses a pre-trained high capacity network to train a much smaller network, suitable for edge devices. In this paper, for the first time, we study the applicability and challenges of using KD for time-series data for wearable devices. Successful application of KD requires specific choices of data augmentation methods during training. However, it is not yet known if there exists a coherent strategy for choosing an augmentation approach during KD. In this paper, we report the results of a detailed study that compares and contrasts various common choices and some hybrid data augmentation strategies in KD based human activity analysis. Research in this area is often limited as there are not many comprehensive databases available in the public domain from wearable devices. Our study considers databases from small scale publicly available to one derived from a large scale interventional study into human activity and sedentary behavior. We find that the choice of data augmentation techniques during KD have a variable level of impact on end performance, and find that the optimal network choice as well as data augmentation strategies are specific to a dataset at hand. However, we also conclude with a general set of recommendations that can provide a strong baseline performance across databases. △ Less

Submitted 31 December, 2021; originally announced January 2022.

Three-year annual modulation search with COSINE-100

Authors: COSINE-100 Collaboration, :, G. Adhikari, E. Barbosa de Souza, N. Carlin, J. J. Choi, S. Choi, A. C. Ezeribe, L. E. França, C. Ha, I. S. Hahn, S. J. Hollick, E. J. Jeon, J. H. Jo, H. W. Joo, W. G. Kang, M. Kauer, H. Kim, H. J. Kim, J. Kim, K. W. Kim, S. H. Kim, S. K. Kim, W. K. Kim, Y. D. Kim , et al. (34 additional authors not shown)

Abstract: COSINE-100 is a direct detection dark matter experiment that aims to test DAMA/LIBRA's claim of dark matter discovery by searching for a dark matter-induced annual modulation signal with NaI(Tl) detectors. We present new constraints on the annual modulation signal from a dataset with a 2.82 yr livetime utilizing an active mass of 61.3 kg, for a total exposure of 173 kg$\cdot$yr. This new result fe… ▽ More COSINE-100 is a direct detection dark matter experiment that aims to test DAMA/LIBRA's claim of dark matter discovery by searching for a dark matter-induced annual modulation signal with NaI(Tl) detectors. We present new constraints on the annual modulation signal from a dataset with a 2.82 yr livetime utilizing an active mass of 61.3 kg, for a total exposure of 173 kg$\cdot$yr. This new result features an improved event selection that allows for both lowering the energy threshold to 1 keV and a more precise time-dependent background model. In the 1-6 keV and 2-6 keV energy intervals, we observe best-fit values for the modulation amplitude of 0.0067$\pm$0.0042 and 0.0051$\pm$0.0047 counts/(day$\cdot$kg$\cdot$keV), respectively, with a phase fixed at 152.5 days. △ Less

Submitted 28 October, 2022; v1 submitted 16 November, 2021; originally announced November 2021.

Comments: 11 pages, 8 figures

Journal ref: Phys. Rev. D 106, 052005 (2022)

A novel measurement of initial-state gluon radiation in hadron collisions using Drell-Yan events

Authors: CDF Collaboration, T. Aaltonen, S. Amerio, D. Amidei, A. Anastassov, A. Annovi, J. Antos, G. Apollinari, J. A. Appel, T. Arisawa, A. Artikov, J. Asaadi, W. Ashmanskas, B. Auerbach, A. Aurisano, F. Azfar, W. Badgett, T. Bae, A. Barbaro-Galtieri, V. E. Barnes, B. A. Barnett, P. Barria, P. Bartos, M. Bauce, F. Bedeschi , et al. (375 additional authors not shown)

Abstract: A study of initial-state gluon radiation (ISR) in hadron collisions is presented using Drell-Yan (DY) events produced in proton-antiproton collisions by the Tevatron collider at a center-of-mass energy of 1.96 TeV. This paper adopts a novel approach which uses the mean value of the Z/$γ^*$ transverse momentum $<p_T^{DY}>$ in DY events as a powerful observable to characterize the effect of ISR. In… ▽ More A study of initial-state gluon radiation (ISR) in hadron collisions is presented using Drell-Yan (DY) events produced in proton-antiproton collisions by the Tevatron collider at a center-of-mass energy of 1.96 TeV. This paper adopts a novel approach which uses the mean value of the Z/$γ^*$ transverse momentum $<p_T^{DY}>$ in DY events as a powerful observable to characterize the effect of ISR. In a data sample corresponding to an integrated luminosity of 9.4 fb$^{-1}$ collected with the CDF Run II detector, $<p_T^{DY}>$ is measured as a function of the Z/$γ^*$ invariant mass. It is found that these two observables have a dependence, $<p_T^{DY}> = -8 + 2.2 \ln m_{DY}^2$ [GeV/c], where $m_{DY}$ is the value of the Z/$γ^*$ mass measured in units of GeV/$c^2$. This linear dependence is observed for the first time in this analysis. It may be exploited to model the effect of ISR and constrain its impact in other processes. △ Less

Submitted 28 October, 2021; v1 submitted 28 October, 2021; originally announced October 2021.

Comments: 14 pages, 14 figures

Searching for low-mass dark matter via Migdal effect in COSINE-100

Authors: G. Adhikari, N. Carlin, J. J. Choi, S. Choi, A. C. Ezeribe, L. E. Franca, C. Ha, I. S. Hahn, S. J. Hollick, E. J. Jeon, J. H. Jo, H. W. Joo, W. G. Kang, M. Kauer, H. Kim, H. J. Kim, J. Kim, K. W. Kim, S. H. Kim, S. K. Kim, W. K. Kim, Y. D. Kim, Y. H. Kim, Y. J. Ko, H. J. Kwon , et al. (31 additional authors not shown)

Abstract: We report on the search for weakly interacting massive particle (WIMP) dark matter candidates in the galactic halo that interact with sodium and iodine nuclei in the COSINE-100 experiment and produce energetic electrons that accompany recoil nuclei via the the Migdal effect. The WIMP mass sensitivity of previous COSINE-100 searches that relied on the detection of ionization signals produced by tar… ▽ More We report on the search for weakly interacting massive particle (WIMP) dark matter candidates in the galactic halo that interact with sodium and iodine nuclei in the COSINE-100 experiment and produce energetic electrons that accompany recoil nuclei via the the Migdal effect. The WIMP mass sensitivity of previous COSINE-100 searches that relied on the detection of ionization signals produced by target nuclei recoiling from elastic WIMP-nucleus scattering was restricted to WIMP masses above $\sim$5 GeV/$c^2$ by the detectors' 1 keVee energy-electron-equivalent threshold. The search reported here looks for recoil signals enhanced by the Migdal electrons that are ejected during the scattering process. This is particularly effective for the detection of low-mass WIMP scattering from the crystals' sodium nuclei in which a relatively larger fraction of the WIMP's energy is transferred to the nucleus recoil energy and the excitation of its orbital electrons. In this analysis, the low-mass WIMP search window of the COSINE-100 experiment is extended to WIMP mass down to 200 MeV/$c^2$. The low-mass WIMP sensitivity will be further improved by lowering the analysis threshold based on a multivariable analysis technique. We consider the influence of these improvements and recent developments in detector performance to re-evaluate sensitivities for the future COSINE-200 experiment. With a 0.2 keVee analysis threshold and high light-yield NaI(Tl) detectors (22 photoelectrons/keVee), the COSINE-200 experiment can explore low-mass WIMPs down to 20 MeV/$c^2$ and probe previously unexplored regions of parameter space. △ Less

Submitted 10 January, 2022; v1 submitted 12 October, 2021; originally announced October 2021.

Comments: 10 pages, 8 figures

Journal ref: Phys. Rev. D 105, 042006 (2022)

FA-GAN: Feature-Aware GAN for Text to Image Synthesis

Authors: Eunyeong Jeon, Kunhee Kim, Dai** Kim

Abstract: Text-to-image synthesis aims to generate a photo-realistic image from a given natural language description. Previous works have made significant progress with Generative Adversarial Networks (GANs). Nonetheless, it is still hard to generate intact objects or clear textures (Fig 1). To address this issue, we propose Feature-Aware Generative Adversarial Network (FA-GAN) to synthesize a high-quality… ▽ More Text-to-image synthesis aims to generate a photo-realistic image from a given natural language description. Previous works have made significant progress with Generative Adversarial Networks (GANs). Nonetheless, it is still hard to generate intact objects or clear textures (Fig 1). To address this issue, we propose Feature-Aware Generative Adversarial Network (FA-GAN) to synthesize a high-quality image by integrating two techniques: a self-supervised discriminator and a feature-aware loss. First, we design a self-supervised discriminator with an auxiliary decoder so that the discriminator can extract better representation. Secondly, we introduce a feature-aware loss to provide the generator more direct supervision by employing the feature representation from the self-supervised discriminator. Experiments on the MS-COCO dataset show that our proposed method significantly advances the state-of-the-art FID score from 28.92 to 24.58. △ Less

Submitted 2 September, 2021; originally announced September 2021.

Comments: ICIP 2021

Alpha backgrounds in the AMoRE-Pilot experiment

Authors: V. Alenkov, H. W. Bae, J. Beyer, R. S. Boiko, K. Boonin, O. Buzanov, N. Chanthima, M. K. Cheoun, S. H. Choi, F. A. Danevich, M. Djamal, D. Drung, C. Enss, A. Fleischmann, A. Gangapshev, L. Gastaldo, Yu. M. Gavriljuk, A. Gezhaev, V. D. Grigoryeva, V. Gurentsov, D. H. Ha, C. Ha, E. J. Ha, I. Hahn, E. J. Jeon , et al. (81 additional authors not shown)

Abstract: The Advanced Mo-based Rare process Experiment (AMoRE)-Pilot experiment is an initial phase of the AMoRE search for neutrinoless double beta decay of $^{100}$Mo, with the purpose of investigating the level and sources of backgrounds. Searches for neutrinoless double beta decay generally require ultimately low backgrounds. Surface $α$ decays on the crystals themselves or nearby materials can deposit… ▽ More The Advanced Mo-based Rare process Experiment (AMoRE)-Pilot experiment is an initial phase of the AMoRE search for neutrinoless double beta decay of $^{100}$Mo, with the purpose of investigating the level and sources of backgrounds. Searches for neutrinoless double beta decay generally require ultimately low backgrounds. Surface $α$ decays on the crystals themselves or nearby materials can deposit a continuum of energies that can be as high as the $Q$-value of the decay itself and may fall in the region of interest (ROI). To understand these background events, we studied backgrounds from radioactive contaminations internal to and on the surface of the crystals or nearby materials with Geant4-based Monte Carlo simulations. In this study, we report on the measured $α$ energy spectra fitted with the corresponding simulated spectra for six crystal detectors, where sources of background contributions could be identified through high energy $α$ peaks and continuum parts in the energy spectrum for both internal and surface contaminations. We determine the low-energy contributions from internal and surface $α$ contaminations by extrapolating from the $α$ background fitting model. △ Less

Submitted 5 December, 2022; v1 submitted 16 July, 2021; originally announced July 2021.

The environmental monitoring system at the COSINE-100 experiment

Authors: H. Kim, G. Adhikari, E. Barbosa de Souza, N. Carlin, J. J. Choi, S. Choi, M. Djamal, A. C. Ezeribe, L. E. França, C. Ha, I. S. Hahn, E. J. Jeon, J. H. Jo, H. W. Joo, W. G. Kang, M. Kauer, H. J. Kim, K. W. Kim, S. H. Kim, S. K. Kim, W. K. Kim, Y. D. Kim, Y. H. Kim, Y. J. Ko, E. K. Lee , et al. (28 additional authors not shown)

Abstract: The COSINE-100 experiment is designed to test the DAMA experiment which claimed an observation of a dark matter signal from an annual modulation in their residual event rate. To measure the 1 %-level signal amplitude, it is crucial to control and monitor nearly all environmental quantities that might systematically mimic the signal. The environmental monitoring also helps ensure a stable operation… ▽ More The COSINE-100 experiment is designed to test the DAMA experiment which claimed an observation of a dark matter signal from an annual modulation in their residual event rate. To measure the 1 %-level signal amplitude, it is crucial to control and monitor nearly all environmental quantities that might systematically mimic the signal. The environmental monitoring also helps ensure a stable operation of the experiment. Here, we describe the design and performance of the centralized environmental monitoring system for the COSINE-100 experiment. △ Less

Submitted 28 November, 2021; v1 submitted 15 July, 2021; originally announced July 2021.

Comments: 13 pages, 11 figures

Measurement of the charge asymmetry of electrons from the decays of $W$ bosons produced in $p\bar{p}$ collisions at $\sqrt{s}=1.96$ TeV

Authors: CDF Collaboration, T. Aaltonen, S. Amerio, D. Amidei, A. Anastassov, A. Annovi, J. Antos, G. Apollinari, J. A. Appel, T. Arisawa, A. Artikov, J. Asaadi, W. Ashmanskas, B. Auerbach, A. Aurisano, F. Azfar, W. Badgett, T. Bae, A. Barbaro-Galtieri, V. E. Barnes, B. A. Barnett, P. Barria, P. Bartos, M. Bauce, F. Bedeschi , et al. (376 additional authors not shown)

Abstract: At the Fermilab Tevatron proton-antiproton ($p\bar{p}$) collider, high-mass electron-neutrino ($eν$) pairs are produced predominantly in the process $p \bar{p} \rightarrow W(\rightarrow eν) + X$. The asymmetry of the electron and positron yield as a function of their pseudorapidity constrain the slope of the ratio of the $u$- to $d$-quark parton distributions versus the fraction of the proton mome… ▽ More At the Fermilab Tevatron proton-antiproton ($p\bar{p}$) collider, high-mass electron-neutrino ($eν$) pairs are produced predominantly in the process $p \bar{p} \rightarrow W(\rightarrow eν) + X$. The asymmetry of the electron and positron yield as a function of their pseudorapidity constrain the slope of the ratio of the $u$- to $d$-quark parton distributions versus the fraction of the proton momentum carried by the quarks. This paper reports on the measurement of the electron-charge asymmetry using the full data set recorded by the Collider Detector at Fermilab in 2001--2011 and corresponding to 9.1~fb$^{-1}$ of integrated luminosity. The measurement significantly improves the precision of the Tevatron constraints on the parton-distribution functions of the proton. Numerical tables of the measurement are provided. △ Less

Submitted 2 November, 2021; v1 submitted 9 July, 2021; originally announced July 2021.

Comments: 27 pages, 25 figures. To be published in PRD

Report number: FERMILAB-PUB-21-293-E

SEEN: Sharpening Explanations for Graph Neural Networks using Explanations from Neighborhoods

Authors: Hyeoncheol Cho, Youngrock Oh, Eunjoo Jeon

Abstract: Explaining the foundations for predictions obtained from graph neural networks (GNNs) is critical for credible use of GNN models for real-world problems. Owing to the rapid growth of GNN applications, recent progress in explaining predictions from GNNs, such as sensitivity analysis, perturbation methods, and attribution methods, showed great opportunities and possibilities for explaining GNN predi… ▽ More Explaining the foundations for predictions obtained from graph neural networks (GNNs) is critical for credible use of GNN models for real-world problems. Owing to the rapid growth of GNN applications, recent progress in explaining predictions from GNNs, such as sensitivity analysis, perturbation methods, and attribution methods, showed great opportunities and possibilities for explaining GNN predictions. In this study, we propose a method to improve the explanation quality of node classification tasks that can be applied in a post hoc manner through aggregation of auxiliary explanations from important neighboring nodes, named SEEN. Applying SEEN does not require modification of a graph and can be used with diverse explainability techniques due to its independent mechanism. Experiments on matching motif-participating nodes from a given graph show great improvement in explanation accuracy of up to 12.71% and demonstrate the correlation between the auxiliary explanations and the enhanced explanation accuracy through leveraging their contributions. SEEN provides a simple but effective method to enhance the explanation quality of GNN model outputs, and this method is applicable in combination with most explainability techniques. △ Less

Submitted 15 June, 2021; originally announced June 2021.

Comments: 15 pages, 3 figures

Fine-Grained Attention for Weakly Supervised Object Localization

Authors: Junghyo Sohn, Eun** Jeon, Wonsik Jung, Eunsong Kang, Heung-Il Suk

Abstract: Although recent advances in deep learning accelerated an improvement in a weakly supervised object localization (WSOL) task, there are still challenges to identify the entire body of an object, rather than only discriminative parts. In this paper, we propose a novel residual fine-grained attention (RFGA) module that autonomously excites the less activated regions of an object by utilizing informat… ▽ More Although recent advances in deep learning accelerated an improvement in a weakly supervised object localization (WSOL) task, there are still challenges to identify the entire body of an object, rather than only discriminative parts. In this paper, we propose a novel residual fine-grained attention (RFGA) module that autonomously excites the less activated regions of an object by utilizing information distributed over channels and locations within feature maps in combination with a residual operation. To be specific, we devise a series of mechanisms of triple-view attention representation, attention expansion, and feature calibration. Unlike other attention-based WSOL methods that learn a coarse attention map, having the same values across elements in feature maps, our proposed RFGA learns fine-grained values in an attention map by assigning different attention values for each of the elements. We validated the superiority of our proposed RFGA module by comparing it with the recent methods in the literature over three datasets. Further, we analyzed the effect of each mechanism in our RFGA and visualized attention maps to get insights. △ Less

Submitted 11 April, 2021; originally announced April 2021.

Comments: 16 pages, 11 figures

Strong constraints from COSINE-100 on the DAMA dark matter results using the same sodium iodide target

Authors: G. Adhikari, E. Barbosa de Souza, N. Carlin, J. J. Choi, S. Choi, M. Djamal, A. C. Ezeribe, L. E. França, C. Ha, I. S. Hahn, E. J. Jeon, J. H. Jo, H. W. Joo, W. G. Kang, M. Kauer, H. Kim, H. J. Kim, K. W. Kim, S. H. Kim, S. K. Kim, W. K. Kim, Y. D. Kim, Y. H. Kim, Y. J. Ko, E. K. Lee , et al. (28 additional authors not shown)

Abstract: We present new constraints on dark matter interactions using 1.7 years of COSINE-100 data. The COSINE-100 experiment, consisting of 106 kg of tallium-doped sodium iodide (NaI(Tl)) target material, is aimed at testing DAMA's claim of dark matter observation using the same NaI(Tl) detectors. Improved event selection requirements, a more precise understanding of the detector background and the use of… ▽ More We present new constraints on dark matter interactions using 1.7 years of COSINE-100 data. The COSINE-100 experiment, consisting of 106 kg of tallium-doped sodium iodide (NaI(Tl)) target material, is aimed at testing DAMA's claim of dark matter observation using the same NaI(Tl) detectors. Improved event selection requirements, a more precise understanding of the detector background and the use of a larger data set considerably enhances the COSINE-100 sensitivity for dark matter detection. No signal consistent with the dark matter interaction is identified, and rules out model-dependent dark matter interpretations of the DAMA signals in the specific context of standard halo model with the same NaI(Tl) target for various interaction hypotheses. △ Less

Submitted 26 August, 2021; v1 submitted 8 April, 2021; originally announced April 2021.

Journal ref: Sci. Adv. 7, eabk2699 (2021)

Privacy-Preserving and Sustainable Contact Tracing Using Batteryless Bluetooth Low-Energy Beacons

Authors: Pietro Tedeschi, Kang Eun Jeon, James She, Simon Wong, Spiridon Bakiras, Roberto Di Pietro

Abstract: Contact tracing is the techno-choice of reference to address the COVID-19 pandemic. Many of the current approaches have severe privacy and security issues and fail to offer a sustainable contact tracing infrastructure. We address these issues introducing an innovative, privacy-preserving, sustainable, and experimentally tested architecture that leverages batteryless BLE beacons. Contact tracing is the techno-choice of reference to address the COVID-19 pandemic. Many of the current approaches have severe privacy and security issues and fail to offer a sustainable contact tracing infrastructure. We address these issues introducing an innovative, privacy-preserving, sustainable, and experimentally tested architecture that leverages batteryless BLE beacons. △ Less

Submitted 21 December, 2021; v1 submitted 10 March, 2021; originally announced March 2021.

Comments: 8 pages

Identification of new isomers in $^{228}$Ac : Impact on dark matter searches

Authors: K. W. Kim, G. Adhikari, E. Barbosa de Souza, N. Carlin, J. J. Choi, S. Choi, M. Djamal, A. C. Ezeribe, L. E. Franca, C. Ha, I. S. Hahn, E. J. Jeon, J. H. Jo, H. W. Joo, W. G. Kang, M. Kauer, H. Kim, H. J. Kim, S. H. Kim, S. K. Kim, W. K. Kim, Y. D. Kim, Y. H. Kim, Y. J. Ko, E. K. Lee , et al. (28 additional authors not shown)

Abstract: We report the identification of metastable isomeric states of $^{228}$Ac at 6.28 keV, 6.67 keV and 20.19 keV, with lifetimes of an order of 100 ns. These states are produced by the $β$-decay of $^{228}$Ra, a component of the $^{232}$Th decay chain, with $β$ Q-values of 39.52 keV, 39.13 keV and 25.61 keV, respectively. Due to its low Q-value as well as the relative abundance of $^{232}$Th and their… ▽ More We report the identification of metastable isomeric states of $^{228}$Ac at 6.28 keV, 6.67 keV and 20.19 keV, with lifetimes of an order of 100 ns. These states are produced by the $β$-decay of $^{228}$Ra, a component of the $^{232}$Th decay chain, with $β$ Q-values of 39.52 keV, 39.13 keV and 25.61 keV, respectively. Due to its low Q-value as well as the relative abundance of $^{232}$Th and their progeny in low background experiments, these observations potentially impact the low-energy background modeling of dark matter search experiments. △ Less

Submitted 12 August, 2021; v1 submitted 3 March, 2021; originally announced March 2021.

Comments: 9 pages, 9 figures

Journal ref: Eur. Phys. J. C 81 (2021) 746

Background modeling for dark matter search with 1.7 years of COSINE-100 data

Authors: G. Adhikari, P. Adhikari, E. Barbosa de Souza, N. Carlin, J. J. Choi, S. Choi, M. Djamal, A. C. Ezeribe, L. E. Franca, C. Ha, I. S. Hahn, E. J. Jeon, J. H. Jo, W. G. Kang, M. Kauer, G. S. Kim, H. Kim, H. J. Kim, K. W. Kim, N. Y. Kim, S. K. Kim, Y. D. Kim, Y. H. Kim, Y. J. Ko, V. A. Kudryavtsev , et al. (25 additional authors not shown)

Abstract: We present a background model for dark matter searches using an array of NaI(Tl) crystals in the COSINE-100 experiment that is located in the Yangyang underground laboratory. The model includes background contributions from both internal and external sources, including cosmogenic radionuclides and surface $^{210}$Pb contamination. To build the model in the low energy region, with a threshold of 1… ▽ More We present a background model for dark matter searches using an array of NaI(Tl) crystals in the COSINE-100 experiment that is located in the Yangyang underground laboratory. The model includes background contributions from both internal and external sources, including cosmogenic radionuclides and surface $^{210}$Pb contamination. To build the model in the low energy region, with a threshold of 1 keV, we used a depth profile of $^{210}$Pb contamination in the surface of the NaI(Tl) crystals determined in a comparison between measured and simulated spectra. We also considered the effect of the energy scale errors propagated from the statistical uncertainties and the nonlinear detector response at low energies. The 1.7 years COSINE-100 data taken between October 21, 2016 and July 18, 2018 were used for this analysis. Our Monte Carlo simulation provides a non-Gaussian peak around 50 keV originating from beta decays of bulk $^{210}$Pb in a good agreement with the measured background. This model estimates that the activities of bulk $^{210}$Pb and $^{3}$H are dominating the background rate that amounts to an average level of 2.85$\pm$0.15 counts/day/keV/kg in the energy region of (1-6) keV, using COSINE-100 data with a total exposure of 97.7 kg$\cdot$years. △ Less

Submitted 6 September, 2022; v1 submitted 27 January, 2021; originally announced January 2021.

Search for sterile neutrino oscillation using RENO and NEOS data

Authors: Z. Atif, J. H. Choi, B. Y. Han, C. H. Jang, H. I. Jang, J. S. Jang, E. J. Jeon, S. H. Jeon, K. K. Joo, K. Ju, D. E. Jung, H. J. Kim, H. S. Kim, J. G. Kim, J. H. Kim, B. R. Kim, J. Y. Kim, J. Y. Kim, S. B. Kim, S. Y. Kim, W. Kim, Y. D. Kim, Y. J. Ko, E. Kwon, D. H. Lee , et al. (22 additional authors not shown)

Abstract: We present a reactor model independent search for sterile neutrino oscillation using 2\,509\,days of RENO near detector data and 180 days of NEOS data. The reactor related systematic uncertainties are significantly suppressed as both detectors are located at the same reactor complex of Hanbit Nuclear Power Plant. The search is performed by electron antineutrino\,($\overlineν_e$) disappearance betw… ▽ More We present a reactor model independent search for sterile neutrino oscillation using 2\,509\,days of RENO near detector data and 180 days of NEOS data. The reactor related systematic uncertainties are significantly suppressed as both detectors are located at the same reactor complex of Hanbit Nuclear Power Plant. The search is performed by electron antineutrino\,($\overlineν_e$) disappearance between six reactors and two detectors with baselines of 294\,m\,(RENO) and 24\,m\,(NEOS). A spectral comparison of the NEOS prompt-energy spectrum with a no-oscillation prediction from the RENO measurement can explore reactor $\overlineν_e$ oscillations to sterile neutrino. Based on the comparison, we obtain a 95\% C.L. excluded region of $0.1<|Δm_{41}^2|<7$\,eV$^2$. We also obtain a 68\% C.L. allowed region with the best fit of $|Δm_{41}^2|=2.41\,\pm\,0.03\,$\,eV$^2$ and $\sin^2 2θ_{14}$=0.08$\,\pm\,$0.03 with a p-value of 8.2\%. Comparisons of obtained reactor antineutrino spectra at reactor sources are made among RENO, NEOS, and Daya Bay to find a possible spectral variation. △ Less

Submitted 6 September, 2022; v1 submitted 2 November, 2020; originally announced November 2020.

Comments: 6 pages, 5 figures: This manuscript has been significantly revised by the joint reanalysis by RENO and NEOS Collaborations. (In the previous edition, the RENO collaboration used publicly available NEOS data to evaluate the expected neutrino spectrum at NEOS.); auxiliary file for data release including error matrix (RENO_NEOS_data_release.txt)

Journal ref: Phys. Rev. D 105, L111101 (2022)

Pulse Shape Discrimination of Fast Neutron Background using Convolutional Neural Network for NEOS II

Authors: NEOS II Collaboration, Y. Jeong, B. Y. Han, E. J. Jeon, H. S. Jo, D. K. Kim, J. Y. Kim, J. G. Kim, Y. D. Kim, Y. J. Ko, H. M. Lee, M. H. Lee, J. Lee, C. S. Moon, Y. M. Oh, H. K. Park, K. S. Park, S. H. Seo, K. Siyeon, G. M. Sun, Y. S. Yoon, I. Yu

Abstract: Pulse shape discrimination plays a key role in improving the signal-to-background ratio in NEOS analysis by removing fast neutrons. Identifying particles by looking at the tail of the waveform has been an effective and plausible approach for pulse shape discrimination, but has the limitation in sorting low energy particles. As a good alternative, the convolutional neural network can scan the entir… ▽ More Pulse shape discrimination plays a key role in improving the signal-to-background ratio in NEOS analysis by removing fast neutrons. Identifying particles by looking at the tail of the waveform has been an effective and plausible approach for pulse shape discrimination, but has the limitation in sorting low energy particles. As a good alternative, the convolutional neural network can scan the entire waveform as they are to recognize the characteristics of the pulse and perform shape classification of NEOS data. This network provides a powerful identification tool for all energy ranges and helps to search unprecedented phenomena of low-energy, a few MeV or less, neutrinos. △ Less

Submitted 28 September, 2020; originally announced September 2020.

Comments: 7 figures

A Novel RL-assisted Deep Learning Framework for Task-informative Signals Selection and Classification for Spontaneous BCIs

Authors: Wonjun Ko, Eun** Jeon, Heung-Il Suk

Abstract: In this work, we formulate the problem of estimating and selecting task-relevant temporal signal segments from a single EEG trial in the form of a Markov decision process and propose a novel reinforcement-learning mechanism that can be combined with the existing deep-learning based BCI methods. To be specific, we devise an actor-critic network such that an agent can determine which timepoints need… ▽ More In this work, we formulate the problem of estimating and selecting task-relevant temporal signal segments from a single EEG trial in the form of a Markov decision process and propose a novel reinforcement-learning mechanism that can be combined with the existing deep-learning based BCI methods. To be specific, we devise an actor-critic network such that an agent can determine which timepoints need to be used (informative) or discarded (uninformative) in composing the intention-related features in a given trial, and thus enhancing the intention identification performance. To validate the effectiveness of our proposed method, we conducted experiments with a publicly available big MI dataset and applied our novel mechanism to various recent deep-learning architectures designed for MI classification. Based on the exhaustive experiments, we observed that our proposed method helped achieve statistically significant improvements in performance. △ Less

Submitted 30 June, 2020; originally announced July 2020.

Comments: 8 pages, 6 figures, 2 tables, and under review

ACM Class: I.2.8

Lowering the energy threshold in COSINE-100 dark matter searches

Authors: G. Adhikari, E. Barbosa de Souza, N. Carlin, J. J. Choi, S. Choi, M. Djamal, A. C. Ezeribe, L. E. Franca, C. Ha, I. S. Hahn, E. J. Jeon, J. H. Jo, W. G. Kang, M. Kauer, H. Kim, H. J. Kim, K. W. Kim, S. K. Kim, Y. D. Kim, Y. H. Kim, Y. J. Ko, E. K. Lee, H. S. Lee, J. Lee, J. Y. Lee , et al. (21 additional authors not shown)

Abstract: COSINE-100 is a dark matter detection experiment that uses NaI(Tl) crystal detectors operating at the Yangyang underground laboratory in Korea since September 2016. Its main goal is to test the annual modulation observed by the DAMA/LIBRA experiment with the same target medium. Recently DAMA/LIBRA has released data with an energy threshold lowered to 1 keV, and the persistent annual modulation beh… ▽ More COSINE-100 is a dark matter detection experiment that uses NaI(Tl) crystal detectors operating at the Yangyang underground laboratory in Korea since September 2016. Its main goal is to test the annual modulation observed by the DAMA/LIBRA experiment with the same target medium. Recently DAMA/LIBRA has released data with an energy threshold lowered to 1 keV, and the persistent annual modulation behavior is still observed at 9.5$σ$. By lowering the energy threshold for electron recoils to 1 keV, COSINE-100 annual modulation results can be compared to those of DAMA/LIBRA in a model-independent way. Additionally, the event selection methods provide an access to a few to sub-GeV dark matter particles using constant rate studies. In this article, we discuss the COSINE-100 event selection algorithm, its validation, and efficiencies near the threshold. △ Less

Submitted 21 March, 2021; v1 submitted 28 May, 2020; originally announced May 2020.

Measurement of the cosmic muon annual and diurnal flux variation with the COSINE-100 detector

Authors: COSINE-100 Collaboration, :, H. Prihtiadi, G. Adhikari, E. Barbosa de Souza, N. Carlin, J. J. Choi, S. Choi, M. Djamal, A. C. Ezeribe, L. E. França, C. Ha, I. S. Hahn, E. J. Jeon, J. H. Jo, W. G. Kang, M. Kauer, H. Kim, H. J. Kim, K. W. Kim, S. K. Kim, Y. D. Kim, Y. H. Kim, Y. J. Ko, E. K. Lee , et al. (23 additional authors not shown)

Abstract: We report measurements of annual and diurnal modulations of the cosmic-ray muon rate in the Yangyang underground laboratory (Y2L) using 952 days of COSINE-100 data acquired between September 2016 and July 2019. A correlation of the muon rate with the atmospheric temperature is observed and its amplitude on the muon rate is determined. The effective atmospheric temperature and muon rate variations… ▽ More We report measurements of annual and diurnal modulations of the cosmic-ray muon rate in the Yangyang underground laboratory (Y2L) using 952 days of COSINE-100 data acquired between September 2016 and July 2019. A correlation of the muon rate with the atmospheric temperature is observed and its amplitude on the muon rate is determined. The effective atmospheric temperature and muon rate variations are positively correlated with a measured effective temperature coefficient of $α_{T}$ = 0.80 $\pm$ 0.11. This result is consistent with a model of meson production in the atmosphere. We also searched for a diurnal modulation in the underground muon rate by comparing one-hour intervals. No significant diurnal modulation of the muon rate was observed. △ Less

Submitted 28 May, 2020; v1 submitted 27 May, 2020; originally announced May 2020.

Comments: 12 pages, 9 figures

Development of ultra-pure NaI(Tl) detectors for the COSINE-200 experiment

Authors: B. J. Park, J. J. Choe, J. S. Choi, O. Gileva, C. Ha, A. Iltis, E. J. Jeon, D. Y. Kim, K. W. Kim, S. K. Kim, Y. D. Kim, Y. J. Ko, C. H. Lee, H. S. Lee, I. S. Lee, M. H. Lee, S. H. Lee, S. J. Ra, J. K. Son, K. A. Shin

Abstract: The annual modulation signal observed by the DAMA experiment is a long-standing question in the community of dark matter direct detection. This necessitates an independent verification of its existence using the same detection technique. The COSINE-100 experiment has been operating with 106~kg of low-background NaI(Tl) detectors providing interesting checks on the DAMA signal. However, due to high… ▽ More The annual modulation signal observed by the DAMA experiment is a long-standing question in the community of dark matter direct detection. This necessitates an independent verification of its existence using the same detection technique. The COSINE-100 experiment has been operating with 106~kg of low-background NaI(Tl) detectors providing interesting checks on the DAMA signal. However, due to higher backgrounds in the NaI(Tl) crystals used in COSINE-100 relative to those used for DAMA, it was difficult to reach final conclusions. Since the start of COSINE-100 data taking in 2016, we also have initiated a program to develop ultra-pure NaI(Tl) crystals for COSINE-200, the next phase of the experiment. The program includes efforts of raw powder purification, ultra-pure NaI(Tl) crystal growth, and detector assembly techniques. After extensive research and development of NaI(Tl) crystal growth, we have successfully grown a few small-size (0.61$-$0.78 kg) thallium-doped crystals with high radio-purity. A high light yield has been achieved by improvements of our detector assembly technique. Here we report the ultra-pure NaI(Tl) detector developments at the Institute for Basic Science, Korea. The technique developed here will be applied to the production of NaI(Tl) detectors for the COSINE-200 experiment. △ Less

Submitted 31 August, 2020; v1 submitted 13 April, 2020; originally announced April 2020.

Comments: 12 pages, 12 figures

Journal ref: Eur. Phys. J. C 80, 814 (2020)

The COSINE-100 Liquid Scintillator Veto System

Authors: G. Adhikari, E. Barbosa de Souza, N. Carlin, J. J. Choi, S. Choi, M. Djamal, A. C. Ezeribe, L. E. Franca, C. Ha, I. S. Hahn, E. J. Jeon, J. H. Jo, W. G. Kang, M. Kauer, H. Kim, H. J. Kim, K. W. Kim, S. K. Kim, Y. D. Kim, Y. H. Kim, Y. J. Ko, E. K. Lee, H. S. Lee, J. Lee, J. Y. Lee , et al. (21 additional authors not shown)

Abstract: This paper describes the liquid scintillator veto system for the COSINE-100 dark matter experiment and its performance. The COSINE-100 detector consists of eight NaI(Tl) crystals immersed in 2200~L of linear alkylbenzene-based liquid scintillator. The liquid scintillator tags between 65 and 75\% of the internal $^{40}$K background in the 2--6 keV energy region. We also describe the background mode… ▽ More This paper describes the liquid scintillator veto system for the COSINE-100 dark matter experiment and its performance. The COSINE-100 detector consists of eight NaI(Tl) crystals immersed in 2200~L of linear alkylbenzene-based liquid scintillator. The liquid scintillator tags between 65 and 75\% of the internal $^{40}$K background in the 2--6 keV energy region. We also describe the background model for the liquid scintillator, which is primarily used to assess its energy calibration and threshold. △ Less

Submitted 14 May, 2021; v1 submitted 5 April, 2020; originally announced April 2020.

Journal ref: Nucl. Instrum. Methods A 1006, 165431 (2021)

Multi-Scale Neural network for EEG Representation Learning in BCI

Authors: Wonjun Ko, Eun** Jeon, Seungwoo Jeong, Heung-Il Suk

Abstract: Recent advances in deep learning have had a methodological and practical impact on brain-computer interface research. Among the various deep network architectures, convolutional neural networks have been well suited for spatio-spectral-temporal electroencephalogram signal representation learning. Most of the existing CNN-based methods described in the literature extract features at a sequential le… ▽ More Recent advances in deep learning have had a methodological and practical impact on brain-computer interface research. Among the various deep network architectures, convolutional neural networks have been well suited for spatio-spectral-temporal electroencephalogram signal representation learning. Most of the existing CNN-based methods described in the literature extract features at a sequential level of abstraction with repetitive nonlinear operations and involve densely connected layers for classification. However, studies in neurophysiology have revealed that EEG signals carry information in different ranges of frequency components. To better reflect these multi-frequency properties in EEGs, we propose a novel deep multi-scale neural network that discovers feature representations in multiple frequency/time ranges and extracts relationships among electrodes, i.e., spatial representations, for subject intention/condition identification. Furthermore, by completely representing EEG signals with spatio-spectral-temporal information, the proposed method can be utilized for diverse paradigms in both active and passive BCIs, contrary to existing methods that are primarily focused on single-paradigm BCIs. To demonstrate the validity of our proposed method, we conducted experiments on various paradigms of active/passive BCI datasets. Our experimental results demonstrated that the proposed method achieved performance improvements when judged against comparable state-of-the-art methods. Additionally, we analyzed the proposed method using different techniques, such as PSD curves and relevance score inspection to validate the multi-scale EEG signal information capturing ability, activation pattern maps for investigating the learned spatial filters, and t-SNE plotting for visualizing represented features. Finally, we also demonstrated our method's application to real-world problems. △ Less

Submitted 1 March, 2020; originally announced March 2020.

Comments: 11 pages, 7 figures